Optionally skip individual torsions when QC step fails

[mattwthompson]

In #292 a user described an issue in which as few as one of the structures in a torsion drive failed the QC generation step¹, causing an error that stopped the entire process.

The proposed solution is to skip particular torsions when something like this fails, but not error out the whole fit. This could be exposed to the user as an option, possibly including finer-grained options such as trying again with hydrogen-containing bonds constrained.

Footnotes

1. In this case, there was a proton transfer during QC optimization, probably resulting from a highly rotated torsion pushing the molecule into a structure it otherwise wouldn't have accessed. This was observed in a zwitterion with the charged groups fairly far apart from each other. ↩